1. Field
The present invention relates generally to web-based contests and ranking systems, and in one particular example, to ranking multiple contest entries in a web-based contest via preference ballots.
2. Related Art
Various web-based voting systems are known. For instance, approval voting is known and generally characterized by a user voting for entries the user likes. Generally, the entry with the greatest number of votes is deemed the winner by such systems. In practice, particularly where there are many entries and a direct ballot can be cast for someone, the contest turns into a pure popularity contest. Such a voting system often suffers from a “clone” problem, where similar candidates may “split” votes. For example, if there are three entries and two of them are similar (e.g., would appeal similarly to users), the two similar entries are at risk of splitting votes because they likely appeal to the same sets of people. Accordingly, the third entry, which might be ranked third by those users voting on the two similar entries may receive the most votes and win.
Other know voting systems include “star” voting or “range voting,” which is generally characterized by allowing voters to ascribe a number to an entry (e.g., a rating between 1 and 5). There are several ways of tallying such votes, e.g., as averaging the number of star ratings for each entry; the entry with the highest mean rating winning. Star voting often suffers from issues of factoring in the number of votes for each entry, as it will typically vary widely between entries, thereby making it difficult to rank entries.
Additionally, several pair-wise voting system are known for matching-up two items for display to a user and receiving a user entered vote for one of the two items. The items may then be ranked based on winning percentages (e.g., the highest or lowest rates of winning in a match-up). One such exemplary known system includes the Kittenwar web site (kittenwar.com), where users may upload photos to be voted on by other users, vote on uploaded photos, and view the “winningest” and “losingest” kitten photos based on their winning percentage in pair-wise match-ups. For example, during a voting process, two pictures of kittens are displayed and voted on by a user selection, e.g., clicking on one of the pictures. After voting on the displayed photos, two new photos are displayed and voted on by the user, and so on.
Ranking entries based on winning percentage in such pair-wise voting systems, however, may suffer in instances where some entries have been involved in votes more frequently than others (and are thus more accurately tested against the pool of entries). Additionally, as contest entries increase in numbers it may become increasingly difficult for voters to view all of the entries.